Hub
ComponentComponent

Document Similarity Predictor

Last edited: 

The Document Similarity Predictor applies the model obtained by the Document Similarity Learner to a test document. It computes the cosine similarity between the original corpus of documents table and the test document.

Component details

Input ports
  1. Type: Table
    Corpus of Documents
    An input table containing the original corpus with the related document vectors.
  2. Type: DocumentVectorPortObject
    Document Vector Model
    A model containing node settings as well as column names of the term feature space.
  3. Type: Table
    Test Document
    An input table containing the new test document.
Output ports
  1. Type: Table
    Selected Similar Documents
    A table containing the selected similar documents
  2. Type: Flow Variable
    Count of Most Similar Documents per Input Document
    A single variable set to the count of matching documents per input document